Cost-effective prescribing: trying to hit the target in Ontario and Australia
- Bernie J. O'Brien
- Aust Prescr 2002;25:128-30
- 1 December 2002
- DOI: 10.18773/austprescr.2002.125
The Canadian province of Ontario does not subsidise prescription drugs for all of its citizens. Despite serving fewer beneficiaries, the Ontario system is facing the same financial pressures as the Australian Pharmaceutical Benefits Scheme. Both systems are using similar strategies to encourage the cost-effective use of drugs. Some drugs can only be prescribed for specific indications and others require the approval of the government before they can be prescribed. Ontario recently tried to limit its expenditure on new drugs to the costs forecast by the manufacturers. The outcome of this controversial policy is not yet known, but it emphasises the need for accurate information about prescribing patterns.
In Canada and Australia expenditure on prescription drugs is growing. The government of Ontario in Canada annually spends close to Can$2 billion of taxpayers' money on prescription drugs. This is the equivalent of A$2.25 billion (A$1 = Can$0.88). As in Australia, an evaluation system has been established to ensure that medicines are used where they are most cost-effective. The Ontario experience has some lessons for and from Australia. Both countries are wrestling with the same problem: of designing a system that effectively guides prescribers to treat patients cost-effectively, yet maintains an appropriate degree of clinical freedom.
Canada has a comprehensive national system of universal public health insurance for medical services similar to Australia. Unlike Australia, out-of-hospital prescription medicines are not covered by the national system and are considered a fiscal responsibility for each province. Consequently, in provinces such as Ontario there are multiple payers for drugs. For example, employed people commonly have prescription drug coverage as an employment benefit although they would share some of the costs. The public payer for drugs in Ontario is the Ontario Drug Benefits Program. This covers about 18% of the population of the province. The primary beneficiaries are those aged over 65 years and people with a specific catastrophic illness or low income.
In 2000 - 01 the Ontario Drug Benefits Program had 49 million prescription claims from its 2.08 million beneficiaries for a total government cost of Can$1.9 billion. The majority of expenditure (67%) is for elderly people. A small percentage of claimants (5%) have annual claims over Can$3000 and account for 27% of all drug costs. The three largest categories of drug expenditures are cardiovascular (Can$422 million), antilipidaemic (Can$226 million) and gastrointestinal (Can$200 million). The 'top ten' drugs in terms of expenditures (Table 1) are similar to the top 10 drugs prescribed in Australia.1 For example, in 2000 - 01 the lipid lowering drug atorvastatin was number one in Ontario (Can$87 million) and number two in Australia (A$280 million).
A major concern in Ontario is the increasing rate of growth of expenditure. In 2000 - 01 annual expenditure grew by 15% (Can$248 million) compared with only 2% in 1992 - 93. The introduction of 10 new products in 2000 - 01 accounted for 70% of expenditure growth. A significant impact (Can$45 million) resulted from the introduction of celecoxib and rofecoxib for treatment of arthritis.
Like the Pharmaceutical Benefits Advisory Committee in Australia, Ontario has a committee which advises the Minister of Health. This Drug Quality and Therapeutics Committee (DQTC) consists of 10 physicians and two pharmacists. It meets monthly to consider submissions made by pharmaceutical manufacturers for the listing of their products on the Ontario formulary. In addition to data on a drug's effectiveness and safety, the manufacturer is required to provide evidence of cost-effectiveness or 'value for money'. Guidelines were published in 1994 on the required form, content and conduct of such economic analyses.3 Members of an economic subcommittee of DQTC carry out expert technical reviews of the economic analyses in the submissions.
The DQTC has several options open to it when recommending a drug for reimbursement. A drug can be listed on the formulary as a 'general benefit' which means it can be prescribed without restriction by any licensed medical practitioner. At the other extreme, a so-called 'Section 8' reimbursement means that the physician must make a written application to the Ministry of Health to justify the need to use the restricted drug. For example, the osteoporosis drug alendronate is a Section 8 benefit; the cost will only be reimbursed if the doctor documents that their patient has 'failed' therapy (e.g. poor efficacy or tolerability) with etidronate. In 2000 - 01 there were 2466 requests for reimbursement for alendronate under Section 8 of which 75% were approved at a cost of Can$788,000.
Between a general benefit and Section 8 is a category of reimbursement which is expanding rapidly. This 'limited use' category is a form of restricted reimbursement that requests the physician to prescribe the drug for patients meeting defined clinical criteria. The key difference between limited use and Section 8 is that it is simply an 'honour system' which trusts the physicians to follow prescribing guidance. There are many examples of limited use drugs, but the most recent debates have been about celecoxib and rofecoxib. Physicians are asked to only prescribe these drugs for patients with arthritis who have an increased risk of gastrointestinal bleeding because it is in these patients that the drugs are most beneficial and cost-effective.
Placing celecoxib and rofecoxib on the Ontario formulary under the limited use category exposes the government to financial risk if prescribers do not abide by the honour system and ignore the limited use criteria. For the government, the 'nightmare scenario' is that the aches and sprains adequately managed with cheap anti-inflammatory drugs get switched to more costly new drugs. In the Australian context I have heard this phenomenon referred to as 'leakage'; once a drug is subsidised for a specified indication and patient group, usage can 'leak' into other patient groups where the drug is less cost-effective. The risk of leakage raises questions of measurement and management. How can a government payer create systems for monitoring appropriate drug use and how should the risk of leakage enter into negotiations with manufacturers?
There are two ways in which drug utilisation review can be used to support limited use criteria. The first is using aggregate or patient-level administrative claims data to monitor trends in drug usage, substitution and other health care usage following formulary listing. For example, the extent to which celecoxib and rofecoxib will lead to reduced prescribing of gastro-protective drugs such as misoprostol is a component of cost-effectiveness models and will be watched keenly. The second method is the use of 'real-time' prescription advice and/or adjudication for reimbursement using office-based electronic medical records. The electronic medical record holds great promise for precisely determining a patient's eligibility for a limited use medicine, but it clearly poses some threats, both to the clinical freedom of prescribers and to the privacy of patients.
As part of the submission for listing provided to DQTC a manufacturer must make a forecast of how much of the drug will be prescribed over the next three years and how much this will cost. This forecast is known as a 'budget impact analysis' and the chief executive officer of each company, prior to listing, must provide a signed letter to the Ministry of Health declaring this forecast.
The forecast of drug expenditure has become a crucial part of the submission because the Ministry of Health has changed its approach to expenditure risk management. In an initial stance - which totally 'blindsided' the industry - the Ministry announced that it would only pay for a new medicine up to the expenditure forecasted in the submission. Faced with a storm of protest on this risk-shifting policy, the Ministry softened its position somewhat and established the Drug Utilization Advisory Committee as an advisory board on circumstances where a manufacturer 'overshoots' their forecast expenditure. It is too early to know how the Drug Utilization Advisory Committee will work and so the 'penalty' for overshooting the forecast remains unclear.
These recent policy developments on agreed expenditure envelopes have some important strategic implications for manufacturers making submissions. Essentially a manufacturer is now entering into a price-volume agreement with the government where it can control the price but has less than 100% control over utilisation once the drug is in the hands of prescribers. The risky business decision for the company is where to set its forecast expenditure for the drug, given two important unknowns: the precise extent of utilisation and the potential penalty for an overshoot in expenditure. It is also a game of strategy for the government which must decide to accept or reject the listing of a drug based on both the cost-effectiveness data and the uncertain forecast expenditure.
This paper was written while I was on sabbatical leave at the University of Sydney and I am grateful to my hosts at the Centre for Health Economics Research and Evaluation and the NHMRC Clinical Trials Centre.
The Ontario Drug Benefits Program homepage
The ODBP formulary
Ontario guidelines on economic analysis for drug submissions
Dr O'Brien is a member of the economic subcommittee of the Ontario Drug Quality and Therapeutics Committee.
Department of Clinical Epidemiology and Biostatistics, McMaster University and Centre for Evaluation of Medicines, St Joseph's Hospital, Ontario, Canada